GoPro3+ on test with standard Phantom

[DennyR]

The final conclusion is that the Phantom really needs a third pan axis. This 2.7 K footage has no post stab. but shows some unwanted pan movements when not in perfectly calm conditions. The only thing changed was the saturation which was given a slight increase as vimeo tends to flatten the colors in compression. I went out and shot everything in one day and chose difficult lighting to see how the camera coped. At the end I was shooting at the beach bar at around midnight and only then could I see a little noise in the dark areas. The distortion still is a problem but I have a solution that corrects this and can deliver perfect linear correction to photogrammetric standards, but this involves using a template in photoshop that creates a batch process through the targa files. So the 3+ is a definite step up.:tennis:

 

[Stacky]

There are threads all over the web with users complaining about the focus point being 3-4 feet from the lens but everything in the distance is soft. The 3+ may have improved one thing but the numbers of complaints from new owners of the 3+ indicate they have created another problem elsewhere.

 

[DennyR]

Stacky
I have not seen that problem yet, I have a second camera that I will try today and see if I can find any grounds for that conclusion. Some people screw-up with everything you give them so I would take that with a pinch of salt. The Cineform Studio software has some interesting features, one of which can cause some weird distortion around the center. It's called H-DYNAMIC and should be used with caution. IT IS NOT A CORRECTION FILTER FOR FISH EYE. I tested it on a couple of those clips and it shows strange artifacts at the center. It also has a sharpness feature which seems to work well but on my 15" retina display it looks awesome straight out of the camera.

What these guys are probably seeing is motion blur caused by too much camera movement which looks worse the further out you look, it could also be caused by using too soft rubber mounts. I used the hardest set (40deg.) and also have a loosish tie wrap around them. The softer you make it then the worse will be the vibration seen by the camera. With a properly prepared machine you don't need any rubbers. The only other deviation from a bog standard Phantom was the use of a great lipo from MaxAmps taking it from 2200 to 2800 mah. with a slight increase in weight.
This particular Phantom had good motor balance but I have seen some that were really bad. The Phantom has to be the easiest machine to fly straight out of the box. I tried sitting the model at 45 deg. on the ground and it still lifted off straight up.

I am about to test it in a Caipirinha at higher levels.

 

[Stacky]

Denny from what I have been able to gather it sounds like GoPro when putting in a better lens also decided to adjust the focus so that things were sharper closer in. Because the GoPro is designed more for action sports and not really for AP that makes sense to me, you want the optimum sharpness to be where the action is. It was after all designed for action sports. From what I can gather GoPro have made an effort to reduce the chromatic abheration with the new lens . It does sound like GoPro are listening so it will be interesting if in the future they produce a GoPro with the ability to adjust the focus slightly. Essentially we are using an action sports camera for AP.

 

[DennyR]

That makes sense, Chromatic Aberration is generally more common in larger lenses for larger sensors especially when wide open. Being smaller makes it easier to get the focussing over a large depth of field and control Chromatic Aberration. I expect that the Sharpness feature in Cineform was put there to help.

As I can correct the distortion in Photoshop I cant see why they could not introduce a camera feature that electronically corrects this digitally in camera.

Maybe when we get to Gp4......

 

[twistedshifter]

Compared my GP3 with my GP3+ side by side. There is no doubt a focusing issue with the 3+ (focus is set too close out of the box.) I decided to take apart the 3+ and unscrew the lens about 1mm to reset the focus for greater depth of field. Problem fixed if you are brave enough to pull apart and mod your new GP3+.

 

[Stacky]

From what i can gather once people have refocussed the 3+ produces a pretty nice file.

 

[jes1111]

Without wishing to provoke a scuffle here, I just don't "get" the popularity of the GoPro as an AP camera. It's a sports/action cam! Why start with a cheapo 170 degree fisheye lens (that's set to focus close-in!) on a tiny sensor in a minuscule body and then worry about "the distortion", "the noise" and "the file"? What did you expect?

The GoPro promo videos look great because a) they're shot by great photographers, b) they're graded and edited by highly talented individuals and c) the shots are always appropriate to the camera's raison d'etre - i.e. close-in, high-energy ACTION. Apart from a small handful of sports sequences, I've yet to see an AP video from a GoPro that doesn't make me yawn and cringe at the same time. :dejection:

Nevertheless, thank you Denny for demonstrating that the GP3+ is still a crappy little camera for AP

 

[Stacky]

Without wishing to provoke a scuffle here, I just don't "get" the popularity of the GoPro as an AP camera. It's a sports/action cam! Why start with a cheapo 170 degree fisheye lens (that's set to focus close-in!) on a tiny sensor in a minuscule body and then worry about "the distortion", "the noise" and "the file"? What did you expect?

The GoPro promo videos look great because a) they're shot by great photographers, b) they're graded and edited by highly talented individuals and c) the shots are always appropriate to the camera's raison d'etre - i.e. close-in, high-energy ACTION. Apart from a small handful of sports sequences, I've yet to see an AP video from a GoPro that doesn't make me yawn and cringe at the same time. :dejection:

Nevertheless, thank you Denny for demonstrating that the GP3+ is still a crappy little camera for AP

How can you not understand why the gopro is popular for ap hobbyists?

 

[DennyR]

I have just finished the filtering template to not only correct the fish eye but corrected it to photogrammetric standards. I have also almost finished shooting the Girne Harbor at night section for the XXXXXXX. You may be a little surprised at the sharpness, color, stability and the ability to see in darkness. Shot at 2.7 K and cineform. I also created the camera moves alone. You will be hard pressed to show me something shot with a DSLR that even comes close. I fly low over people and the sea for extended periods and at also 2500 feet.. Why... because I can and with total confidence. Not something that I would advise you to try with anything remotely heavy lift. It is amazing what some people can do with a crappy little camera and a machine that weighs 1 kilo. The workflow to create this is known about in pro circles but to learn how to do it you will need to get one of my tutorial videos which are coming soon.

I could have used my TDR and a Red Epic but because I wanted dynamic content that nobody else could do, I used the right tool for the job.

BTW Having just been sold a duff 3 axis Alexmos by Viacopter ( jussi doesn't have the common decency to reply) I bought a much better product from Virtual Robotix all I can say is WOW.

 

[jes1111]

I agree that you don't need a Red to create a good video, of course. But is there really such a difference in weight/duration/safety with, say, a NEX/GH (or a BMCC) rig? Wouldn't the PP workflow surely be a whole lot easier/faster and, most importantly, of predictable output quality? I don't doubt that your Girne Harbor piece will be stunning - not doubting your abilities at all - and I look forward to seeing it. But, from your description of the process, I still don't see how the GoPro is definitively "the right tool for the job". I would certainly accept your argument if the choice were between a "mirrorless" camera and a DSLR or RED.

You've reminded me: I emailed VR to buy one of their boards but never got a reply - must chase them up. The spec/features certainly look tasty. I'm also intrigued by John Cunningham's new stuff (after a long absence).

 

[DennyR]

The whole point is that you can do everything that you will ever be asked to shoot with a small quad like a Phantom and a GoPro with a 3 axis gimbal. Whether you create the resolution with the glass or with software matters not as far as the client is concerned. If it looks right then it is right. What is not right is when one cannot accept that it is highly illogical to haul male jewelry into the sky in the misguided belief that you will make better productions by doing so. Everything will work against you. As soon as you put a mirror on a DSLR then the film plane moves out and the lens design is compromised along with all of the extra weight. The Nex-7 was tried here along with several other cameras. in the end it comes down to the simple fact that these cameras have too many bells and whistles that you don't need. The associated weight penalty is massive against a GoPro and those cameras don't shoot 4K. or any of the other good features like 240 fps. if you know how to rectify the image then it is no contest. Just seen the latest s800 evo at my local dealer. It weighs over 6.5 kilos and fly's for 8 mins. with 2x 1100 batteries. And all that just to lift a Nex-7...

I now consider that anything over 1.5 kilos is a waste of time because you are immediately restricted in where and how you can fly, that could mean that your opposition will get the job.
As always I have had motor gliders for high level work that is often used as an establishing shot. the TBS wings are a good choice.

 

[jes1111]

It's a little difficult to keep up with your recommendations, Denny. You've built a succession of weird and wonderful machines, declaring each of them (at the time) as rendering everything else invalid. Now all cameras bigger or more expensive than a GoPro are "male jewellery"? And SLR lenses are a compromised design?

 

[DennyR]

I have dozens of models that were designed to do specific jobs, some of those jobs were thought to be impossible. That is what I call R&D. It is what makes Red Bull better than Ferrari and Mercedes. It is a reflection of how that company can make and test 500 design changes every week to the same car. Yet only a few of those designs will actually be used in competition. In my work as an aerial pro shooter I have used everything from a Arri 435 to a Red One all the way down to an i-phone. I am not making any recommendations to anyone. I just like to pass on SOME of my observations. If someone enjoys building say a beautiful Kratos or a DW then that is great that they will get the personal satisfaction from doing that. Modelers like satisfaction from what they build. My concern is that the almost exponential risk of putting heavy in the air will become a factor in the future of our existence. The size of the imaging sensor is what determines the size and weight everything else that you put in the air. As this single component improves then does it not make sense to create vehicles that will also be tailored to that same end. A good example of how the size of the camera affects the weight can be seen in the difference between a Phantom and a S800 EVO, the difference in image quality is disproportional to the weight, cost and risk factor. The nex-7 is not that much bigger than a GoPro yet the final outcome is a vehicle that changes from 1 kilo to 7 kilos.

An example of how AP is developing is that ten years ago aerial survey work was all done on a multi million dollar aircraft that used a 9"x9" roll film camera. it could map the surface of the ground to within about 12 mm. at best. Today Small drones can do that job to an accuracy of 3mm. because they can fly lower and more accurately and thus create better triangulation.

 

[Stacky]

Denny,how do you see the weight of a multi relative to wind as a factor. I have been to Cyprus many years ago and my memory was of frequent sea breezes. My current experience of multis is that in general the bigger heavier machines handle the wind better than a smaller (phantom for example) lighter multi. I live on an Isthmus and so wind and sea breezes are something I have to deal with. With respect to wind I am finding the smaller lighter quads just cant give me footage in a decent breeze that a larger multi can.

 

[DennyR]

I fly in winds of up to 12 m/sec. after that it is possible to fly with smaller props. but if I want good results then I wait for better conditions. The wind here can be very strong at times. I would probably use my TDR if it was essential to get a job done in extreme conditions. What makes the Phantom so stable in light winds is the short arms and small props which produce a fast response. There is a ratio of wind speed against thrust velocity that produces good stability. I don,t have an exact figure for that but looking at the latest S800 EVO they have also come down on the prop loading and are running much faster than before.
I just got back from the North after a good days shooting and I was so surprised at how much tilt could be set on the phantom at take-off and still have it lift straight up without any side movement. Like 45 deg. DJI have definitely got some good code in this instance.